Guides the design and improvement of human annotation programs for machine-learning datasets, covering labeling schemas, acquisition, annotator quality, disagreement, provenance, adjudication, and stopping criteria.
Design and improve human annotation programs for machine-learning training, validation, and evaluation data, including task and label schemas, active acquisition, annotator guidance, calibration, disagreement, adjudication, provenance, and cost-quality stopping. Do not use for general statistical inference, data storage pipelines, model training, or annotation-interface implementation; route those concerns to data-scientist, data-engineering, ml-engineering, or product-design-and-ux.
evals/evals.json
JSON
Other
README.md
Markdown
Other
templates/annotation-guide.md
Markdown
Other
templates/annotation-plan.md
Markdown
Other
templates/quality-report.md
Markdown
Reference
references/acquisition-and-stopping.md
Markdown
Reference
references/annotator-quality.md
Markdown
Reference
references/provenance-and-rights.md
Markdown
Reference
references/source-index.md
Markdown
Reference
references/task-and-label-design.md
Markdown
SKILL.md
SKILL.md
Markdown
Provides sklearn-style estimators for incrementally updating covariance, correlation, precision, and location matrices from streaming observations using constant work per observat…
Enables Python agents to update, assess, compare, and select online covariance, correlation, and precision estimators for streaming or dynamic-universe data.
Wraps positional online covariance estimators to process keyed observations whose named series can remain fixed or change dynamically over time.
Recommends covariance estimator classes for a 2-D dataset using dimension, conditioning, rank, correlation, and tail-behavior features, and exposes those features for inspection.
Guides an AI agent through end-to-end Kaggle competitions, including validation, modeling, OOF predictions, ensembling, remote notebook pipelines, artifact handoffs, submission sc…
Analyzes e-commerce reviews using 22-dimension AI tagging to extract customer profiles, pain points, VOC, product opportunities, anomaly signals, reports, dashboards, and optional…
Assesses, designs, and evolves data architectures through workload analysis, platform and modeling tradeoffs, governance, data products, data mesh, event-driven flows, and migrati…
Queries Open Library's public catalog for books, authors, works, editions, ISBNs, ratings, and cover URLs through a keyless Python CLI.